The rendering of reconstructed images onto an end-device display is of key importance to insure an end-to-end service quality. However, it is not an easy task because of the wide range of colour formats, of capture capability and of display characteristics. Recently, a new and wider colour space format has been proposed by ITU in the document ITU-R Recommendation BT.2020 (known as Rec. 2020) entitled “Parameter values for UHDTV systems for production and international programme exchange” published in April 2012. Consequently, the compatibility with legacy devices has to be considered.
All the rendering devices may not have the capability to adapt to any colour space nor have the required knowledge to perform the optimal colour conversion. Indeed, rather than clipping colours (left part of FIG. 1) one may prefer graceful colour shading for example as illustrated on the right part of FIG. 1. The determination of the colour conversion function is not straightforward because the content creation workflows may include deterministic processing (Colour Space 1 toward Colour Space 2 conversion) but also non-deterministic operations such as colour grading. If two targeted displays with different characteristics are used (e.g. UHDTV display and Rec.709 HDTV display), both the artistic intent and the colour grading may be different. The colour grading depends both on the content characteristics and the reference display.
As depicted on FIG. 2, in video content distribution, a colour transform is usually applied on the decoded pictures so that the mapped decoded pictures are adapted to the end device rendering capability.
This colour transform also known as Colour Mapping Function (CMF) is for example approximated by a 3×3 gain matrix plus an offset (Gain-Offset model) or by a 3D Colour LUT.
There is thus a need to encode a colour transform for example in the form of a 3D Colour LUT in bit-streams, possibly transmitted out-of band. This can provide the necessary flexibility and additional features to applications and services on top of HEVC and SHVC video coding standards.
One solution is to transmit the colour transform or more generally colour metadata at the transport system level in private streams. However, most of the transmission systems discard those metadata because they do not know how to interpret them.